Exim Main Log Huge / Crashing Server

liamw3

Registered
Jun 8, 2007
1
0
151
Evening All,

Wondered if you could help as my host has been useless

Around the same time everynight exim appears to go crazy...

The log file becomes absolutely huge and crashes the server as it eventually uses up ALL 60GB.

I checked an hour ago and the exim main log was at 11GB. Anyway I assumed rotation wasn't working, manually rotated it, and checked cron - all worked fine. I come back an hour later to see it rising in size again, already 1GB...

cat exim_mainlog shows:
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor
2010-04-20 01:11:59 1 accept() failure: Bad file descriptor

cat panic log shows
2010-04-18 15:00:58 IPv4 socket creation failed: Cannot allocate memory
2010-04-18 15:05:14 exim 4.69 daemon started: pid=14003, -q1h, listening for SMTP on port 25 (IPv6 and IPv4) port 26 (IPv6 and IPv4) and for SMTPS on port 465 (IPv6 and IPv4)
2010-04-18 15:05:14 Cannot open main log file "/var/log/exim_mainlog": Permission denied: euid=47 egid=12
2010-04-18 15:05:25 1O3V7l-0004G1-5a <= [email protected] U=root P=local S=525
2010-04-18 15:05:25 1O3V7l-0004G1-5a Cannot open main log file "/var/log/exim_mainlog": Permission denied: euid=47 egid=12
2010-04-19 01:13:13 50 accept() failures: Cannot allocate memory
2010-04-20 01:11:58 50 accept() failures: Cannot allocate memory

So it seems memory related, yet the server works fine throughout the rest of the day until exim decides to go crazy....

[email protected] [/var/log]# ls -lsSh
-bash: start_pipeline: pgrp pipe: Cannot allocate memory
total 2.3G
2.2G -rw-r----- 1 mailnull mail 2.2G Apr 20 01:41 exim_mainlog

The server was responding absolutely fine and throwing no memory errors all day until around this time everynight when it appears the server crashes. As it is filling the exim mainlog with all that rubbish I cannot do anything to the server at the moment... and will have to wait until they restart the VPS from their end.

Any ideas?